Abstract:
This paper evaluates the impact of a decentralisation reform introduced by a new constitution in Kenya on three household financial outcomes: financial inclusion, savings, and borrowing. The reform established 47 independent counties, classified as either 'marginalised counties' or 'privileged counties.' It introduced a formal revenue-sharing system between the central government and these counties, and an equalisation fund designed to address historical injustices and regional inequalities in the marginalised counties. Using difference-in-differences analysis, we found that households in marginalised regions experienced a 25% increase in savings and a 60% increase in borrowing compared to their counterparts in privileged regions. In contrast, the reform led to a 10% decrease in financial inclusion, though this impact was imprecise. Quantile regressions showed that the reform had a greater impact on households at the top of the financial inclusion and savings distribution, while the impact on borrowing was uniform across the distribution. We also investigated the underlying mechanisms that contributed to these effects. Causal mediation analysis revealed that increased budget allocation to marginalised counties contributed more indirectly to the total effect of decentralisation reforms than its direct effect, overruling the impact of the many changes introduced by the constitution. Placebo tests using falsified treated households and treatment times showed no significant effects, suggesting that the observed treatment effect was unlikely to be influenced by confounding factors. These findings contribute to ongoing contentious policy discussions in the country regarding the complete implementation, modification, and duration of the equalisation fund, and they have implications for other countries.